In [1]:
import glob
import time
import cv2
import pickle
import os.path
import random
import scipy.misc
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
from datetime import datetime
from time import sleep
from moviepy.editor import VideoFileClip
from IPython.display import HTML
from mpl_toolkits.mplot3d import Axes3D
from sklearn.svm import SVC
from sklearn.preprocessing import StandardScaler
from sklearn.cross_validation import train_test_split
from sklearn.model_selection import GridSearchCV
from skimage.feature import hog
from scipy.ndimage.measurements import label
%matplotlib inline
Functions for detecting lane line¶
In [2]:
def undistor_image(img, obj_points, img_points, build_pickle=False):
"""
Undistort an image given obj_points, img_points
"""
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
pickle_path = "./camera_cal/wide_dist_pickle.p"
if build_pickle or (not os.path.exists(pickle_path) ):
ret, mtx, dist, rvecs, tvecs = cv2.calibrateCamera(objpoints,
img_points,
(img.shape[1], img.shape[0]),
None, None)
dist_pickle = {}
dist_pickle["mtx"] = mtx
dist_pickle["dist"] = dist
dist_pickle["ret"] = ret
dist_pickle["rvecs"] = rvecs
dist_pickle["tvecs"] = tvecs
pickle.dump( dist_pickle, open( pickle_path, "wb" ) )
return cv2.undistort(gray, mtx, dist, None, mtx)
dist_pickle = pickle.load( open( pickle_path, "rb" ) )
return cv2.undistort(img, dist_pickle["mtx"], dist_pickle["dist"], None, dist_pickle["mtx"])
def debug_images(img1, img2=None, title1="", title2=None):
"""
Debug image by showing a set the titles
"""
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 9))
f.tight_layout()
ax1.imshow(img1, cmap='gray')
ax1.set_title(title1, fontsize=40)
if img2 != None:
ax2.imshow(img2, cmap='gray')
ax2.set_title(title2, fontsize=40)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
def dir_threshold(img, sobel_kernel=3, thresh=(0, np.pi/2)):
"""
Returns binary output of image, filtering out pixels
that are not in the desired direction(tresh)
"""
gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
x = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=sobel_kernel)
y = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=sobel_kernel)
abs_x = np.absolute(x)
abs_y = np.absolute(y)
arc_xy = np.arctan2(abs_y, abs_x)
binary_output = np.zeros_like(arc_xy)
binary_output[(arc_xy >= thresh[0]) & (arc_xy <= thresh[1])] = 1
return binary_output
def abs_sobel_thresh(img, orient='x', sobel_kernel=3, thresh=(0, 255)):
"""
Apply sobel to an image in given direction
returns binary image filtered by thresholds
"""
gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
derivative = (1, 0)
if orient == 'y':
derivative = (0, 1)
sobel = cv2.Sobel(gray, cv2.CV_64F, derivative[0], derivative[1], ksize=sobel_kernel)
abs_sobel = np.absolute(sobel)
scaled = np.uint8(255*abs_sobel/np.max(abs_sobel))
binary_output = np.zeros_like(scaled)
binary_output[(scaled >= thresh[0]) & (scaled <= thresh[1])] = 1
return binary_output
def mag_thresh(img, sobel_kernel=3, mag_thresh=(0, 255)):
"""
Returns binary image, filters out by magnitude
"""
gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
x = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=sobel_kernel)
y = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=sobel_kernel)
mag = np.sqrt(x**2 + y**2)
scaled = np.uint8(mag/(np.max(mag)/255))
binary_output = np.zeros_like(scaled)
binary_output[(scaled >= mag_thresh[0]) & (scaled <= mag_thresh[1])] = 1
return binary_output
def sobel(image, ksize=3, xy_thresh=(20, 255), mag_thresh_=(20, 200), dir_thresh=(0.3, 1.2)):
"""
Returns a filtered binary image combining sobel in x and y
magnitude and direction
"""
gradx = abs_sobel_thresh(image, orient='x', sobel_kernel=ksize, thresh=xy_thresh)
grady = abs_sobel_thresh(image, orient='y', sobel_kernel=ksize, thresh=xy_thresh)
mag_binary = mag_thresh(image, sobel_kernel=ksize, mag_thresh=mag_thresh_)
dir_binary = dir_threshold(image, sobel_kernel=ksize, thresh=dir_thresh)
combined = np.zeros_like(dir_binary)
combined[((gradx == 1) & (grady == 1)) | ((mag_binary == 1) & (dir_binary == 1))] = 1
return combined
def channel_thresh(channel, thresh=(20, 255)):
"""
Returns a filtered channel i.e one the RGB or HSV etc
"""
ch_binary = np.zeros_like(channel)
ch_binary[(channel >= thresh[0]) & (channel <= thresh[1])] = 1
return ch_binary
def color_and_gradient_comb(img, debug=(False, False)):
"""
Returns a combination gradient of image and selected channels
"""
img = np.copy(img)
hls = cv2.cvtColor(img, cv2.COLOR_RGB2HLS)
hsv = cv2.cvtColor(img, cv2.COLOR_RGB2HSV)
s1_channel = hls[:,:,2]
l_channel = hls[:,:,1]
s2_channel = hsv[:,:,1]
v_channel = hsv[:,:,2]
R_channel = img[:,:, 0]
s1_channel_bin = channel_thresh(s1_channel, thresh=(160, 255))
s2_channel_bin = channel_thresh(s2_channel, thresh=(150, 255))
l_channel_bin = channel_thresh(l_channel, thresh=(220, 255))
v_channel_bin = channel_thresh(v_channel, thresh=(110, 255))
R_channel_bin = channel_thresh(R_channel, thresh=(210, 255))
comb_channel = np.zeros_like(s1_channel_bin)
comb_channel[((s1_channel_bin == 1)|
(s2_channel_bin == 1) |
(l_channel_bin == 1)) &
(v_channel_bin == 1) |
(R_channel_bin == 1)] = 1
#sobel on x and y, mag and dir
dxdy_img = sobel(img, ksize=3, xy_thresh=(30, 250), mag_thresh_=(30, 250), dir_thresh=(-1.0, 1.0))
# Stack each channel
# Note color_binary[:, :, 0] is all 0s, effectively an all black image. It might
# be beneficial to replace this channel with something else.
stacked = np.dstack(( np.zeros_like(dxdy_img), dxdy_img, comb_channel))
all_comb = np.zeros_like(dxdy_img)
all_comb[(dxdy_img == 1) | (comb_channel == 1)] = 1
if debug[0] or debug[1]:
#plot original image
if debug[0]:
#plot channels vs binaries
debug_images(s1_channel, s1_channel_bin, 'S1 CH', 'S1 BIN')
debug_images(s2_channel, s2_channel_bin, 'S2 CH', 'S2 BIN')
debug_images(l_channel, l_channel_bin, 'L CH', 'L BIN')
debug_images(v_channel, v_channel_bin, 'V CH', 'V BIN')
debug_images(R_channel, R_channel_bin, 'R CH', 'R BIN')
#plot gradient vs channel comb
if debug[1]:
debug_images(dxdy_img, comb_channel, 'Derivative', 'Channel')
#plot stacked of gradient and channel combinations
debug_images(stacked, all_comb, 'Stacked dx+comb', 'Combined dx+comb')
debug_images(img, img, 'IMG', 'IMG')
return all_comb
def warp(img, debug=False):
"""
Returns a warped image and its inverse Matrix
"""
img = np.copy(img)
shape = img.shape
src = np.float32(
[
[.55*shape[1], .64*shape[0]],
[.45*shape[1], .64*shape[0]],
[.15*shape[1], shape[0]],
[.88*shape[1], shape[0]]
])
_src = src.reshape((-1,1,2)).astype(int)
_input = cv2.polylines(np.copy(img),[_src],True,(255,0,0), 2)
dst = np.float32(
[
[.75*shape[1], 0],
[.25*shape[1], 0],
[.25*shape[1], shape[0]],
[.75*shape[1], shape[0]]
])
M = cv2.getPerspectiveTransform(src, dst)
Minv = cv2.getPerspectiveTransform(dst, src)
warped = cv2.warpPerspective(img, M, (shape[1], shape[0]), flags=cv2.INTER_LINEAR)
_dst = dst.reshape((-1,1,2)).astype(int)
_out_put = cv2.polylines(np.copy(warped),[_dst],True,(255,0,0), 2)
if debug:
debug_images(_input, _out_put, 'input', 'warped')
return (np.uint8(warped), Minv)
def radius(ploty, leftx, rightx, height_m=30.0, width_m=3.7, height_px=720.0, width_px=1280.0):
"""
Compute the radius of a line
"""
y_eval = np.max(ploty)
ym_per_pix = height_m/height_px
xm_per_pix = width_m/width_px
left_fit_cr = np.polyfit(ploty*ym_per_pix, leftx*xm_per_pix, 2)
right_fit_cr = np.polyfit(ploty*ym_per_pix, rightx*xm_per_pix, 2)
left_curverad = ((1 + (2*left_fit_cr[0]*y_eval*ym_per_pix + left_fit_cr[1])**2)**1.5) / np.absolute(2*left_fit_cr[0])
right_curverad = ((1 + (2*right_fit_cr[0]*y_eval*ym_per_pix + right_fit_cr[1])**2)**1.5) / np.absolute(2*right_fit_cr[0])
middle = (leftx[-1] + rightx[-1])//2
carpos = ((width_px//2) - middle)*xm_per_pix
return (left_curverad, right_curverad, carpos)
def plot_lane_area(warped, undist, Minv, ploty, left_fitx, right_fitx, debug=False):
"""
Plots the area between lane lines
"""
warp_zero = np.zeros_like(warped).astype(np.uint8) * 255
color_warp = np.dstack((warp_zero, warp_zero, warp_zero))
pts_left = np.array([np.transpose(np.vstack([left_fitx, ploty]))])
pts_right = np.array([np.flipud(np.transpose(np.vstack([right_fitx, ploty])))])
pts = np.hstack((pts_left, pts_right))
cv2.fillPoly(color_warp, np.int_([pts]), (0,255, 0))
newwarp = cv2.warpPerspective(color_warp, Minv, (undist.shape[1], undist.shape[0]))
result = cv2.addWeighted(undist, 1, newwarp, 0.3, 0)
if debug:
plt.figure()
plt.imshow(result)
plt.figure()
plt.imshow(np.uint8(newwarp))
return result
def find_lane_line(binary_warped, debug=False, boxes=False):
"""
Find lane lines equations base on warped image
"""
histogram = np.sum(binary_warped[int(binary_warped.shape[0]/2):,:], axis=0)
out_img = np.dstack((binary_warped, binary_warped, binary_warped))*255
midpoint = np.int(histogram.shape[0]/2)
leftx_base = np.argmax(histogram[:midpoint])
rightx_base = np.argmax(histogram[midpoint:]) + midpoint
nwindows = 9
window_height = np.int(binary_warped.shape[0]/nwindows)
nonzero = binary_warped.nonzero()
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
leftx_current = leftx_base
rightx_current = rightx_base
margin = 100
minpix = 10
left_lane_inds = []
right_lane_inds = []
for window in range(nwindows):
win_y_low = binary_warped.shape[0] - (window+1)*window_height
win_y_high = binary_warped.shape[0] - window*window_height
win_xleft_low = leftx_current - margin
win_xleft_high = leftx_current + margin
win_xright_low = rightx_current - margin
win_xright_high = rightx_current + margin
if boxes:
cv2.rectangle(out_img,(win_xleft_low,win_y_low),(win_xleft_high,win_y_high),(0,255,0), 2)
cv2.rectangle(out_img,(win_xright_low,win_y_low),(win_xright_high,win_y_high),(0,255,0), 22)
good_left_inds = ((nonzeroy >= win_y_low) & (nonzeroy < win_y_high) & (nonzerox >= win_xleft_low) & (nonzerox < win_xleft_high)).nonzero()[0]
good_right_inds = ((nonzeroy >= win_y_low) & (nonzeroy < win_y_high) & (nonzerox >= win_xright_low) & (nonzerox < win_xright_high)).nonzero()[0]
left_lane_inds.append(good_left_inds)
right_lane_inds.append(good_right_inds)
if len(good_left_inds) > minpix:
leftx_current = np.int(np.mean(nonzerox[good_left_inds]))
if len(good_right_inds) > minpix:
rightx_current = np.int(np.mean(nonzerox[good_right_inds]))
left_lane_inds = np.concatenate(left_lane_inds)
right_lane_inds = np.concatenate(right_lane_inds)
leftx = nonzerox[left_lane_inds]
lefty = nonzeroy[left_lane_inds]
rightx = nonzerox[right_lane_inds]
righty = nonzeroy[right_lane_inds]
left_fit = np.polyfit(lefty, leftx, 2)
right_fit = np.polyfit(righty, rightx, 2)
ploty = np.linspace(0, binary_warped.shape[0]-1, binary_warped.shape[0] )
left_fitx = left_fit[0]*ploty**2 + left_fit[1]*ploty + left_fit[2]
right_fitx = right_fit[0]*ploty**2 + right_fit[1]*ploty + right_fit[2]
out_img[nonzeroy[left_lane_inds], nonzerox[left_lane_inds]] = [255, 0, 0]
out_img[nonzeroy[right_lane_inds], nonzerox[right_lane_inds]] = [0, 0, 255]
if debug:
plt.figure()
plt.imshow(out_img)
plt.plot(left_fitx, ploty, color='yellow', markersize=12)
plt.plot(right_fitx, ploty, color='yellow', markersize=12)
plt.xlim(0, 1280)
plt.ylim(720, 0)
data = {
"out_img": out_img,
"left_fit": left_fit,
"right_fit": right_fit,
"left_fitx": left_fitx,
"right_fitx": right_fitx,
"ploty": ploty
}
return data
def find_second_frame(binary_warped, left_fit, right_fit, debug=False):
"""
Find lane lines of next frame base on last left and right fits
"""
nonzero = binary_warped.nonzero()
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
margin = 100
left_lane_inds = ((nonzerox > (left_fit[0]*(nonzeroy**2) + left_fit[1]*nonzeroy + left_fit[2] - margin)) & (nonzerox < (left_fit[0]*(nonzeroy**2) + left_fit[1]*nonzeroy + left_fit[2] + margin)))
right_lane_inds = ((nonzerox > (right_fit[0]*(nonzeroy**2) + right_fit[1]*nonzeroy + right_fit[2] - margin)) & (nonzerox < (right_fit[0]*(nonzeroy**2) + right_fit[1]*nonzeroy + right_fit[2] + margin)))
leftx = nonzerox[left_lane_inds]
lefty = nonzeroy[left_lane_inds]
rightx = nonzerox[right_lane_inds]
righty = nonzeroy[right_lane_inds]
left_fit = np.polyfit(lefty, leftx, 2)
right_fit = np.polyfit(righty, rightx, 2)
ploty = np.linspace(0, binary_warped.shape[0]-1, binary_warped.shape[0] )
left_fitx = left_fit[0]*ploty**2 + left_fit[1]*ploty + left_fit[2]
right_fitx = right_fit[0]*ploty**2 + right_fit[1]*ploty + right_fit[2]
out_img = np.dstack((binary_warped, binary_warped, binary_warped))*255
window_img = np.zeros_like(out_img)
out_img[nonzeroy[left_lane_inds], nonzerox[left_lane_inds]] = [255, 0, 0]
out_img[nonzeroy[right_lane_inds], nonzerox[right_lane_inds]] = [0, 0, 255]
left_line_window1 = np.array([np.transpose(np.vstack([left_fitx-margin, ploty]))])
left_line_window2 = np.array([np.flipud(np.transpose(np.vstack([left_fitx+margin, ploty])))])
left_line_pts = np.hstack((left_line_window1, left_line_window2))
right_line_window1 = np.array([np.transpose(np.vstack([right_fitx-margin, ploty]))])
right_line_window2 = np.array([np.flipud(np.transpose(np.vstack([right_fitx+margin, ploty])))])
right_line_pts = np.hstack((right_line_window1, right_line_window2))
if debug:
cv2.fillPoly(window_img, np.int_([left_line_pts]), (0,255, 0))
cv2.fillPoly(window_img, np.int_([right_line_pts]), (0,255, 0))
result = cv2.addWeighted(out_img, 1, window_img, 0.3, 0)
plt.imshow(result)
plt.plot(left_fitx, ploty, color='yellow')
plt.plot(right_fitx, ploty, color='yellow')
plt.xlim(0, 1280)
plt.ylim(720, 0)
data = {
"out_img": out_img,
"left_fit": left_fit,
"right_fit": right_fit,
"left_fitx": left_fitx,
"right_fitx": right_fitx,
"ploty": ploty
}
return data
class Line():
def __init__(self):
self.detected = False
self.recent_xfitted = []
self.bestx = None
self.best_fit = None
self.current_fit = [np.array([False])]
self.line_base_pos = None
self.diffs = np.array([0,0,0], dtype='float')
self.allx = None
self.ally = None
data = None
radius_and_pos = (1,1,1)
def check_first_time(self):
self.detected = True
self.last_data = self.data
self.radius_and_pos = radius(self.data["ploty"],
self.data["left_fitx"],
self.data["right_fitx"],
height_m=30.0, width_m=3.7,
height_px=720.0, width_px=1200.0)
def check(self):
self.radius_and_pos = radius(self.data["ploty"],
self.data["left_fitx"],
self.data["right_fitx"],
height_m=30.0, width_m=3.7,
height_px=720.0, width_px=1200.0)
Section includes vehicle detection functions¶
In [3]:
def convert_color(img, conv='RGB2YCrCb'):
if conv == 'RGB2YCrCb':
return cv2.cvtColor(img, cv2.COLOR_RGB2YCrCb)
if conv == 'BGR2YCrCb':
return cv2.cvtColor(img, cv2.COLOR_BGR2YCrCb)
if conv == 'RGB2LUV':
return cv2.cvtColor(img, cv2.COLOR_RGB2LUV)
if conv == 'RGB2YUV':
return cv2.cvtColor(img, cv2.COLOR_RGB2YUV)
if conv != 'RGB':
if conv == 'HSV':
return cv2.cvtColor(img, cv2.COLOR_RGB2HSV)
elif conv == 'LUV':
return cv2.cvtColor(img, cv2.COLOR_RGB2LUV)
elif conv == 'HLS':
return cv2.cvtColor(img, cv2.COLOR_RGB2HLS)
elif conv == 'YUV':
return cv2.cvtColor(img, cv2.COLOR_RGB2YUV)
elif conv == 'YCrCb':
return cv2.cvtColor(img, cv2.COLOR_RGB2YCrCb)
else: return np.copy(img)
In [4]:
def get_hog_features(img, orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True):
# Call with two outputs if vis==True
if vis == True:
features, hog_image = hog(img, orientations=orient,
pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block),
transform_sqrt=False,
visualise=vis, feature_vector=feature_vec)
return features, hog_image
# Otherwise call with one output
else:
features = hog(img, orientations=orient,
pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block),
transform_sqrt=False,
visualise=vis, feature_vector=feature_vec)
return features
def bin_spatial(img, size=(32, 32)):
color1 = cv2.resize(img[:,:,0], size).ravel()
color2 = cv2.resize(img[:,:,1], size).ravel()
color3 = cv2.resize(img[:,:,2], size).ravel()
return np.hstack((color1, color2, color3))
def color_hist(img, nbins=32, bins_range=(0, 256)):
# Compute the histogram of the color channels separately
channel1_hist = np.histogram(img[:,:,0], bins=nbins, range=bins_range)
channel2_hist = np.histogram(img[:,:,1], bins=nbins, range=bins_range)
channel3_hist = np.histogram(img[:,:,2], bins=nbins, range=bins_range)
# Concatenate the histograms into a single feature vector
hist_features = np.concatenate((channel1_hist[0], channel2_hist[0], channel3_hist[0]))
# Return the individual histograms, bin_centers and feature vector
return hist_features
# Define a function to extract features from a list of images
# Have this function call bin_spatial() and color_hist()
def extract_features(imgs, color_space='RGB', spatial_size=(32, 32),
hist_bins=32, orient=9,
pix_per_cell=8, cell_per_block=2, hog_channel=0,
spatial_feat=True, hist_feat=True, hog_feat=True,
hist_range=(0, 256)):
# Create a list to append feature vectors to
features = []
# Iterate through the list of images
for file in imgs:
file_features = []
# Read in each one by one
image = mpimg.imread(file, format="jpg")
# apply color conversion if other than 'RGB'
feature_image = convert_color(image, conv=color_space)
if spatial_feat == True:
spatial_features = bin_spatial(feature_image, size=spatial_size)
file_features.append(spatial_features)
if hist_feat == True:
# Apply color_hist()
hist_features = color_hist(feature_image, nbins=hist_bins)
file_features.append(hist_features)
if hog_feat == True:
# Call get_hog_features() with vis=False, feature_vec=True
if hog_channel == 'ALL':
hog_features = []
for channel in range(feature_image.shape[2]):
hog_features.append(get_hog_features(feature_image[:,:,channel],
orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True))
hog_features = np.ravel(hog_features)
else:
hog_features = get_hog_features(feature_image[:,:,hog_channel], orient,
pix_per_cell, cell_per_block, vis=False, feature_vec=True)
# Append the new feature vector to the features list
file_features.append(hog_features)
features.append(np.concatenate(file_features))
# Return list of feature vectors
return features
# Define a function to extract features from a single image window
# This function is very similar to extract_features()
# just for a single image rather than list of images
def single_img_features(img, color_space='RGB', spatial_size=(32, 32),
hist_bins=32, orient=9,
pix_per_cell=8, cell_per_block=2, hog_channel=0,
spatial_feat=True, hist_feat=True, hog_feat=True):
#1) Define an empty list to receive features
img_features = []
#2) Apply color conversion if other than 'RGB'
feature_image = convert_color(img, conv=color_space)
#3) Compute spatial features if flag is set
if spatial_feat == True:
spatial_features = bin_spatial(feature_image, size=spatial_size)
#4) Append features to list
img_features.append(spatial_features)
#5) Compute histogram features if flag is set
if hist_feat == True:
hist_features = color_hist(feature_image, nbins=hist_bins)
#6) Append features to list
img_features.append(hist_features)
#7) Compute HOG features if flag is set
if hog_feat == True:
if hog_channel == 'ALL':
hog_features = []
for channel in range(feature_image.shape[2]):
hog_features.extend(get_hog_features(feature_image[:,:,channel],
orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True))
else:
hog_features = get_hog_features(feature_image[:,:,hog_channel], orient,
pix_per_cell, cell_per_block, vis=False, feature_vec=True)
#8) Append features to list
img_features.append(hog_features)
#9) Return concatenated array of features
return np.concatenate(img_features)
In [5]:
# Define a function that takes an image,
# start and stop positions in both x and y,
# window size (x and y dimensions),
# and overlap fraction (for both x and y)
def slide_window(img, x_start_stop=[None, None], y_start_stop=[None, None],
xy_window=(64, 64), xy_overlap=(0.5, 0.5)):
# If x and/or y start/stop positions not defined, set to image size
if x_start_stop[0] == None:
x_start_stop[0] = 0
if x_start_stop[1] == None:
x_start_stop[1] = img.shape[1]
if y_start_stop[0] == None:
y_start_stop[0] = 0
if y_start_stop[1] == None:
y_start_stop[1] = img.shape[0]
# Compute the span of the region to be searched
xspan = x_start_stop[1] - x_start_stop[0]
yspan = y_start_stop[1] - y_start_stop[0]
# Compute the number of pixels per step in x/y
nx_pix_per_step = np.int(xy_window[0]*(1 - xy_overlap[0]))
ny_pix_per_step = np.int(xy_window[1]*(1 - xy_overlap[1]))
# Compute the number of windows in x/y
nx_buffer = np.int(xy_window[0]*(xy_overlap[0]))
ny_buffer = np.int(xy_window[1]*(xy_overlap[1]))
nx_windows = np.int((xspan-nx_buffer)/nx_pix_per_step)
ny_windows = np.int((yspan-ny_buffer)/ny_pix_per_step)
# Initialize a list to append window positions to
window_list = []
# Loop through finding x and y window positions
# Note: you could vectorize this step, but in practice
# you'll be considering windows one by one with your
# classifier, so looping makes sense
for ys in range(ny_windows):
for xs in range(nx_windows):
# Calculate window position
startx = xs*nx_pix_per_step + x_start_stop[0]
endx = startx + xy_window[0]
starty = ys*ny_pix_per_step + y_start_stop[0]
endy = starty + xy_window[1]
# Append window position to list
window_list.append(((startx, starty), (endx, endy)))
# Return the list of windows
return window_list
# Define a function to draw bounding boxes
def draw_boxes(img, bboxes, color=(0, 0, 255), thick=6):
# Make a copy of the image
imcopy = np.copy(img)
# Iterate through the bounding boxes
for bbox in bboxes:
# Draw a rectangle given bbox coordinates
cv2.rectangle(imcopy, bbox[0], bbox[1], color, thick)
# Return the image copy with boxes drawn
return imcopy
# Define a function you will pass an image
# and the list of windows to be searched (output of slide_windows())
def search_windows(img, windows, clf, scaler, color_space='RGB',
spatial_size=(32, 32), hist_bins=32,
hist_range=(0, 256), orient=9,
pix_per_cell=8, cell_per_block=2,
hog_channel=0, spatial_feat=True,
hist_feat=True, hog_feat=True, debug=False):
#1) Create an empty list to receive positive detection windows
on_windows = []
#2) Iterate over all windows in the list
for window in windows:
#3) Extract the test window from original image
test_img = cv2.resize(img[window[0][1]:window[1][1], window[0][0]:window[1][0]], (64, 64))
#4) Extract features for that window using single_img_features()
features = single_img_features(test_img, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat)
#5) Scale extracted features to be fed to classifier
test_features = scaler.transform(np.array(features).reshape(1, -1))
#6) Predict using your classifier
prediction = clf.predict(test_features)
if debug:
debug_images(test_img, title1="window: {}, predict: {}".format(window, prediction))
#7) If positive (prediction == 1) then save the window
if prediction == 1:
on_windows.append(window)
#8) Return windows for positive detections
return on_windows
training classifier to detect vehicles¶
In [6]:
cars = glob.glob("img/train/vehicles/**/*.png")
notcars = glob.glob("img/train/non-vehicles/**/*.png")
random.shuffle(cars)
random.shuffle(cars)
random.shuffle(notcars)
random.shuffle(notcars)
sample_size = 8000
cars = cars[0:sample_size]
notcars = notcars[0:sample_size]
color_space = "YUV" # Can be RGB, HSV, LUV, HLS, YUV, YCrCb
orient = 11 # HOG orientations
pix_per_cell = 16 # HOG pixels per cell
cell_per_block = 2 # HOG cells per block
hog_channel = "ALL" # Can be 0, 1, 2, or "ALL"
spatial_size = (32, 32) # Spatial binning dimensions
hist_bins = 32 # Number of histogram bins
spatial_feat = True # Spatial features on or off
hist_feat = True # Histogram features on or off
hog_feat = True # HOG features on or off
y_start_stop = [450, 656] # Min and max in y to search in slide_window()
car_features = extract_features(cars, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat)
notcar_features = extract_features(notcars, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat)
X = np.vstack((car_features, notcar_features)).astype(np.float64)
X_scaler = StandardScaler().fit(X)
scaled_X = X_scaler.transform(X)
y = np.hstack((np.ones(len(car_features)), np.zeros(len(notcar_features))))
rand_state = np.random.randint(0, 10000)
X_train, X_test, y_train, y_test = train_test_split(
scaled_X, y, test_size=0.4, random_state=rand_state)
#tuned_parameters = [{'kernel': ['sigmoid'], 'gamma': [ 0.0001, 0.00003], 'C': [ 0.5, 1 ]}]
#clf = GridSearchCV(SVC(verbose=True), tuned_parameters, cv=4)
clf = SVC(verbose=True, gamma=0.0001, C=0.5, kernel="rbf")
print('Using:',orient,'orientations',pix_per_cell,
'pixels per cell and', cell_per_block,'cells per block')
print('Feature vector length:', len(X_train[0]))
# Check the training time for the SVC
t=time.time()
clf.fit(X_train, y_train)
t2 = time.time()
print(round(t2-t, 2), 'Seconds to train SVC...')
# Check the score of the SVC
print('Test Accuracy of SVC = ', round(clf.score(X_test, y_test), 4))
# Check the prediction time for a single sample
t=time.time()
#print("Best Parameters")
#print(clf.best_params_)
In [7]:
def add_heat(heatmap, bbox_list):
# Iterate through list of bboxes
for box in bbox_list:
# Add += 1 for all pixels inside each bbox
# Assuming each "box" takes the form ((x1, y1), (x2, y2))
heatmap[box[0][1]:box[1][1], box[0][0]:box[1][0]] += 1
# Return updated heatmap
return heatmap# Iterate through list of bboxes
def apply_threshold(heatmap, threshold):
# Zero out pixels below the threshold
heatmap[heatmap <= threshold] = 0
# Return thresholded map
return heatmap
def draw_labeled_bboxes(img, labels):
# Iterate through all detected cars
for car_number in range(1, labels[1]+1):
# Find pixels with each car_number label value
nonzero = (labels[0] == car_number).nonzero()
# Identify x and y values of those pixels
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
# Define a bounding box based on min/max x and y
bbox = ((np.min(nonzerox), np.min(nonzeroy)), (np.max(nonzerox), np.max(nonzeroy)))
# Draw the box on the image
cv2.rectangle(img, bbox[0], bbox[1], (0,0,255), 6)
# Return the image
return img
def heat_windows(image, box_list, limit, debug=False, heat_flag=False):
image = np.copy(image)
heat = np.zeros_like(image[:,:,0]).astype(np.float)
# Add heat to each box in box list
heat = add_heat(heat,box_list)
# Apply threshold to help remove false positives
heat = apply_threshold(heat,limit)
# Visualize the heatmap when displaying
heatmap = np.clip(heat, 0, 255)
# Find final boxes from heatmap using label function
labels = label(heatmap)
draw_img = draw_labeled_bboxes(np.copy(image), labels)
if debug:
plt.imshow(draw_img)
plt.title('Car Positions')
fig = plt.figure()
plt.imshow(heatmap, cmap='hot')
plt.title('Heat Map')
fig = plt.figure()
if heat_flag:
return draw_img, heatmap, labels
return draw_img
In [8]:
def detect_windows(image, plot=False):
draw_image = np.copy(image)
y_start_stop = [380, 600]
x_start_stop = [540, image.shape[1]]
windows1 = slide_window(image, x_start_stop=x_start_stop, y_start_stop=y_start_stop,
xy_window=(64, 64), xy_overlap=(.5, .5))
y_start_stop = [380, image.shape[0]]
x_start_stop = [240, image.shape[1]]
windows2 = slide_window(image, x_start_stop=x_start_stop, y_start_stop=y_start_stop,
xy_window=(100, 100), xy_overlap=(.6, .8))
y_start_stop = [380, image.shape[0]]
x_start_stop = [240, image.shape[1]]
windows3 = slide_window(image, x_start_stop=x_start_stop, y_start_stop=y_start_stop,
xy_window=(200, 200), xy_overlap=(.6, .6))
windows = windows1 + windows2 + windows3
if plot:
window_img_ = draw_boxes(draw_image, windows, color=(0, 0, 255), thick=6)
plt.imshow(window_img_)
return windows
def find_vehicles(image, clf, not_heat=False):
draw_image = np.copy(image)
windows = detect_windows(image)
hot_windows = search_windows(image, windows, clf, X_scaler, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat, debug=False)
window_img = heat_windows(draw_image, hot_windows, 2)
if not_heat:
window_img = draw_boxes(draw_image, hot_windows, color=(0, 0, 255), thick=6)
return window_img
In [9]:
images_path = glob.glob("./test_images/*.jpg")
tests = [mpimg.imread(path) for path in images_path]
In [10]:
test = tests[3] #mpimg.imread("./test_images/test6.jpg")
for test in tests:
ww=detect_windows(test, plot=True)
plt.figure()
print(len(ww))
In [11]:
debug_images(find_vehicles(tests[0], clf, not_heat=True), find_vehicles(tests[8], clf, not_heat=True), "Boxes", "Boxes")
In [12]:
llabels = []
for test in tests[:6]:
draw_image = np.copy(test)
windows = detect_windows(test)
hot_windows = search_windows(test, windows, clf, X_scaler, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat, debug=False)
bboxed_im = draw_boxes(draw_image, hot_windows, color=(0, 0, 255), thick=6)
_, heat, labels = heat_windows(draw_image, hot_windows, 2, heat_flag=True)
llabels.append(labels[0])
llabels.append(labels[0])
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(10, 5))
f.tight_layout()
ax1.imshow(bboxed_im, cmap='gray')
ax1.set_title("boxes", fontsize=10)
ax2.imshow(heat, cmap='hot')
ax2.set_title("heat map", fontsize=10)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
In [13]:
labels = llabels[0]
for l in llabels:
labels =+ l
plt.imshow(labels, cmap="gray")
Out[13]:
In [14]:
plt.imshow(find_vehicles(tests[8], clf))
Out[14]:
In [15]:
im1 = mpimg.imread(cars[1000], format="jpg")
im2 = mpimg.imread(notcars[1331], format="jpg")
debug_images(im1, im2, "Car", "Not Car")
In [16]:
def plot4(i1, i2, i3, i4):
im1 = i1[0]
t1 = i1[1]
im2 = i2[0]
t2 = i2[1]
im3 = i3[0]
t3 = i3[1]
im4 = i4[0]
t4 = i4[1]
f, (ax1, ax2, ax3, ax4) = plt.subplots(1, 4, figsize=(24, 9))
f.tight_layout()
ax1.imshow(im1, cmap='gray')
ax1.set_title(t1, fontsize=15)
ax2.imshow(im2, cmap='gray')
ax2.set_title(t2, fontsize=15)
ax3.imshow(im3, cmap='gray')
ax3.set_title(t3, fontsize=15)
ax4.imshow(im4, cmap='gray')
ax4.set_title(t4, fontsize=15)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
In [17]:
im1 = mpimg.imread(cars[865], format="jpg")
im2 = mpimg.imread(notcars[865], format="jpg")
im1yuv = cv2.cvtColor(im1, cv2.COLOR_RGB2YUV)
im2yuv = cv2.cvtColor(im2, cv2.COLOR_RGB2YUV)
orient = 11 # HOG orientations
pix_per_cell = 16 # HOG pixels per cell
cell_per_block = 2 # HOG cells per block
hog_channel = "ALL" # Can be 0, 1, 2, or "ALL"
_, im1vis = hog(im1yuv[:,:,0], orientations=orient,
pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block),
transform_sqrt=False,
visualise=True, feature_vector=True)
_, im2vis = hog(im2yuv[:,:,0], orientations=orient,
pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block),
transform_sqrt=False,
visualise=True, feature_vector=True)
plot4((im1[:,:,0], "Car CH-1"),
(im1vis, "Car CH-1 HOG"),
(im2[:,:,0], "Not car CH-1"),
(im2vis, "Not car CH-1 HOG"),
)
plot4((im1[:,:,0], "Car CH-1"),
(im1yuv[:,:,0], "Car CH-1 YUV"),
(im2[:,:,0], "Not car CH-1"),
(im2yuv[:,:,0], "Not car CH-1 YUV"),
)
plot4((im1[:,:,1], "Car CH-2"),
(im1yuv[:,:,1], "Car CH-2 YUV"),
(im2[:,:,1], "Not car CH-2"),
(im2yuv[:,:,1], "Not car CH-2 YUV"),
)
plot4((im1[:,:,2], "Car CH-3"),
(im1yuv[:,:,2], "Car CH-3 YUV"),
(im2[:,:,2], "Not car CH-3"),
(im2yuv[:,:,2], "Not car CH-3 YUV"),
)
In [18]:
corners_pickle = pickle.load( open( "./camera_cal/corners_pickle.p", "rb" ) )
objpoints, imgpoints = (corners_pickle["objpoints"], corners_pickle["imgpoints"])
line = Line()
In [19]:
def process_image(image, debug=False, stacked=True):
"""
Procces each frame of the video and return
The plotted lane lane area over the image.
"""
undistort = undistor_image(image, objpoints, imgpoints, build_pickle=False)
binary = color_and_gradient_comb(undistort, debug=(False, False))
warped, Minv = warp(binary, debug=False)
if line.detected:
line.data = find_second_frame(warped, line.data["left_fit"], line.data["right_fit"], debug=False)
line.check()
else:
line.data = find_lane_line(warped, debug=False)
line.check_first_time()
plotted = plot_lane_area(warped, undistort, Minv, line.data["ploty"], line.data["left_fitx"], line.data["right_fitx"], debug=False)
plotted = find_vehicles(plotted, clf)
# Write Curvature Radius on image
color=[0, 255, 0]
lr, rr, pos = line.radius_and_pos
direction = "left" if pos < 0 else "right"
lcaption = "L[Curvature Radius={0}Km.]".format( round(lr/1000.0, 2) )
rcaption = "R[Curvature Radius={0}Km.]".format( round(rr/1000.0, 2) )
poscaption = "Positon {}m {} of the center".format(round(abs(pos),2), direction )
if not stacked:
cv2.putText(plotted, lcaption,(10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1,color,1,cv2.LINE_AA)
cv2.putText(plotted, rcaption,(10, 70), cv2.FONT_HERSHEY_SIMPLEX, 1,color,1,cv2.LINE_AA)
cv2.putText(plotted, poscaption,(10, 110), cv2.FONT_HERSHEY_SIMPLEX, 1,color,1,cv2.LINE_AA)
#if want to make video with binary or warped images
if stacked:
b = np.uint8(binary)
bbb = np.dstack((b, b, b))*255
w = np.uint8(warped)
www = np.dstack((w, w, w))*255
bbb = cv2.resize(bbb, (426, 240))
www = cv2.resize(www, (426, 240))
fill = cv2.resize(np.zeros_like(bbb), (428, 240) )
cv2.putText(fill, lcaption,(10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.8,color,1,cv2.LINE_AA)
cv2.putText(fill, rcaption,(10, 70), cv2.FONT_HERSHEY_SIMPLEX, 0.8,color,1,cv2.LINE_AA)
cv2.putText(fill, poscaption,(10, 110), cv2.FONT_HERSHEY_SIMPLEX, 0.7,color,1,cv2.LINE_AA)
cv2.putText(fill, "<--- binary gradient + channels",(10, 180), cv2.FONT_HERSHEY_SIMPLEX, 0.7,color,1,cv2.LINE_AA)
cv2.putText(fill, "bird's eye view perspective --->",(10, 200), cv2.FONT_HERSHEY_SIMPLEX, 0.7,color,1,cv2.LINE_AA)
bw = np.concatenate((bbb, fill ,www), axis=1)
return np.concatenate((plotted, bw), axis=0)
return find_vehicles(np.copy(plotted), clf)
In [20]:
clip1 = VideoFileClip("./test_video.mp4")
white_clip = clip1.fl_image(process_image)
%time white_clip.write_videofile('test.mp4', audio=False)
In [21]:
HTML("""
<video width="960" height="540" controls>
<source src="test.mp4">
</video>
""")
Out[21]:
In [22]:
clip1 = VideoFileClip("./project_video.mp4")
white_clip = clip1.fl_image(process_image)
%time white_clip.write_videofile('project_out.mp4', audio=False)
In [23]:
HTML("""
<video width="960" height="540" controls>
<source src="project_out.mp4">
</video>
""")
Out[23]:
In [ ]: